Tangling over Toronto's ozone.

Few actions in high-visibility science fail to stir up debate. A case in point: the report last year by two Canadian scientists that they had detected an increase in the amount of harmful ultraviolet (UV) radiation bathing Toronto.

According to James B. Kerr and C. Thomas McElroy of Environment Canada in Downsview, Ontario, total UV radiation in the Toronto area rose 35 percent per year during winter and 7 percent per year during summer between 1989 and 1993. They attributed the increases in part to a drop in stratospheric ozone concentrations above the city of 4.1 percent in winter and 1.8 percent in summer (SN: 12/4/93, p.382).

Now, three environmental scientists from the University of Virginia in Charlottesville challenge these conclusions. Their reanalysis of the Kerr-McElroy winter data indicates that the Canadian findings stem from "an artifact of the [original] analysis," write Patrick J. Michaels, S. Fred Singer, and Paul C. Knappenberger in the May 27 Science. They assert that the Kerr-McElroy conclusions result from only four UV-B readings (out of 312 used in the original study). "When the four points from March 1993 were eliminated, there was no trend significantly different from zero at any [UV-B] wavelength," the Virginia team says.

In a rebuttal, Kerr and McElroy say the Virginia team erred by using only winter data and by "arbitrarily" rejecting some data. "Perhaps the most serious bias in the analysis by Michaels et al comes from the removal of data points without an apparent physical basis for doing so.... Removal of these critical data points is inappropriate," the two argue.